In clinical studies, the treatment effect may be heterogeneous among patients. It is of interest to identify subpopulations which benefit most from the treatment, regardless of the treatment’s overall performance. In this study, we are interested in subgroup identification in longitudinal studies when nonlinear trajectory patterns are present. Under such a situation, evaluation of the treatment effect entails comparing longitudinal trajectories while subgroup identification requires a further evaluation of differential treatment effects among subgroups induced by moderators. To this end, we propose a tree-structured subgroup identification method, termed “interaction tree for longitudinal trajectories”, which combines mixed effects models with regression splines to model the nonlinear progression patterns among repeated measures. Extensive simulation studies are conducted to evaluate its performance and an application to an alcohol addiction pharmacogenetic trial is presented.
SuXZhouTYanX, et al.
Interaction trees with censored survival data. Int J Biostat2008;
4: 1557–4679.
4.
SuXTsaiCLWangH, et al.
Subgroup analysis via recursive partitioning. J Mach Learn Res2009;
10: 141–158.
5.
DusseldorpEConversanoCVan OsBJ.Combining an additive and tree-based regression model simultaneously: STIMA. J Comput Graph Stat2010;
19: 514–530.
6.
LipkovichIDmitrienkoADenneJ, et al.
Subgroup identification based on differential effect search – A recursive partitioning method for establishing response to treatment in patient subpopulations.Stat Med2011;
30: 2601–2621.
7.
FosterJCTaylorJMRubergSJ.Subgroup identification from randomized clinical trial data.Stat Med2011;
30: 2867–2880.
8.
DooveLLDusseldorpEVan DeunK, et al.
A comparison of five recursive partitioning methods to find person subgroups involved in meaningful treatment–subgroup interactions. Adv Data Anal Classif2014;
8: 403–425.
9.
LipkovichIDmitrienkoAD’AgostinoRBS.Tutorial in biostatistics: data-driven subgroup identification and analysis in clinical trials.Stat Med2017;
36: 136–196.
10.
JohnsonBAAit-DaoudNSeneviratneC, et al.
Pharmacogenetic approach at the serotonin transporter gene as a method of reducing the severity of alcohol drinking.Am J Psychiatry2011;
168: 265–275.
11.
JohnsonBASeneviratneCWangXQ, et al.
Determination of genotype combinations that can predict the outcome of the treatment of alcohol dependence using the 5-ht3 antagonist ondansetron.Am J Psychiatry2013;
170: 1020–1031.
12.
HouJSeneviratneCSuX, et al.
Subgroup identification in personalized treatment of alcohol dependence.Alcohol Clin Exp Res2015;
39: 1253–1259.
13.
SuXMenesesKMcNeesP, et al.
Interaction trees: exploring the differential effects of an intervention programme for breast cancer survivors. J R Stat Soc C Appl Stat2011;
60: 457–474.
14.
ChenJJohnsonBAWangXQ, et al.
Trajectory analyses in alcohol treatment research.Alcohol Clin Exp Res2012;
36: 1442–1448.
15.
AbdolellMLeBlancMStephensD, et al.
Binary partitioning for continuous longitudinal data: categorizing a prognostic variable.Stat Med2002;
21: 3395–3409.
16.
BrandmaierAMVon OertzenTMcArdleJJ, et al.
Structural equation model trees.Psychol Methods2013;
18: 71–86.
17.
StegmannGJacobucciRSerangS, et al.
Recursive partitioning with nonlinear models of change.Multivariate Behav Res2018;
53: 559–570.
18.
FokkemaMSmitsNZeileisA, et al.
Detecting treatment-subgroup interactions in clustered data with generalized linear mixed-effects model trees.Behav Res Methods2018;
50: 2016–2034.
HajjemALarocqueDBellavanceF.Generalized mixed effects regression trees. Stat Probab Lett2017;
126: 114–118.
21.
HajjemABellavanceFLarocqueD.Mixed-effects random forest for clustered data. J Stat Comput Simul2014;
84: 1313–1328.
22.
HajjemABellavanceFLarocqueD.Mixed effects regression trees for clustered data. Stat Probab Lett2011;
81: 451–459.
23.
SelaRJSimonoffJS.Re-em trees: a data mining approach for longitudinal and clustered data. Mach Learn2012;
86: 169–207.
24.
FuWSimonoffJS.Unbiased regression trees for longitudinal and clustered data. Comput Stat Data Anal2015;
88: 53–74.
25.
MurphySA.Optimal dynamic treatment regimes. J R Stat Soc Series B Stat Methodol2003;
65: 331–355.
26.
ChakrabortyBMurphySA.Dynamic treatment regimes.Annu Rev Stat Appl2014;
1: 447–464.
27.
GyörfiLKohlerMKrzyzakA, et al. A distribution-free theory of nonparametric regression.
Berlin, Germany:
Springer Science & Business Media, 2006.
28.
LeBlancMCrowleyJ.Survival trees by goodness of split. J Am Stat Assoc1993;
88: 457–467.
29.
LohWYFuHManM, et al.
Identification of subgroups with differential treatment effects for longitudinal and multiresponse variables.Stat Med2016;
35: 4837–4855.
30.
WilsonEBHilfertyMM.The distribution of chi-square.Proc Natl Acad Sci U S A1931;
17: 684–688.
31.
ChenJLiuLJohnsonBA, et al.
Penalized likelihood estimation for semiparametric mixed models, with application to alcohol treatment research.Stat Med2013;
32: 335–346.
32.
RandWM.Objective criteria for the evaluation of clustering methods. J Am Stat Assoc1971;
66: 846–850.
33.
GatesAJAhnYY.The impact of random models on clustering similarity. J Mach Learn Res2017;
18: 3049–3076.
34.
SchwarzG.Estimating the dimension of a model. Ann Stat1978;
6: 461–464.
35.
AkaikeH.A new look at the statistical model identification. IEEE Trans Automat Contr1974;
19: 716–723.
36.
MokdadAHMarksJSStroupDF, et al.
Actual causes of death in the United States, 2000.JAMA2004;
291: 1238–1245.
37.
KranzlerHRMcKayJR.Personalized treatment of alcohol dependence.Curr Psychiatry Rep2012;
14: 486–493.
38.
HeiligMGoldmanDBerrettiniW, et al.
Pharmacogenetic approaches to the treatment of alcohol addiction.Nat Rev Neurosci2011;
12: 670–684.
39.
SobellLCSobellMB. Timeline follow-back. In: R. Z. Litten et al. (eds.) Measuring alcohol consumption. Berlin, Germany: Springer, 1992, pp.41–72.
40.
LittenRZAllenJP.Measuring alcohol consumption: psychosocial and biochemical methods.
Berlin, Germany:
Springer Science & Business Media, 2012.
41.
KuehnBM.Study suggests gene may predict success of therapies for alcohol dependence.JAMA2011;
305: 984–985.
42.
NiuFZhouJLeTH, et al.
Testing the trajectory difference in a semi-parametric longitudinal model.Stat Methods Med Res2017;
26: 1519–1531.
43.
LanzaSTCoffmanDLXuS.Causal inference in latent class analysis.Struct Equ Modeling2013;
20: 361–383.
44.
LindenA.Using group-based trajectory modelling to enhance causal inference in interrupted time series analysis.J Eval Clin Pract2018;
24: 502–507.
45.
SuXKangJFanJ, et al.
Facilitating score and causal inference trees for large observational studies. J Mach Learn Res2012;
13: 2955–2994.
46.
KangJSuXHitsmanB, et al.
Tree-structured analysis of treatment effects with large observational data. J Appl Stat2012;
39: 513–529.
47.
KangJSuXLiuL, et al.
Causal inference of interaction effects with inverse propensity weighting, g-computation and tree-based standardization. Stat Anal Data Min2014;
7: 323–336.
48.
LohWYZhengW.Regression trees for longitudinal and multiresponse data. Ann Appl Stat2013;
7: 495–522.
49.
BürginRRitschardG.Tree-based varying coefficient regression for longitudinal ordinal responses. Comput Stat Data Anal2015;
86: 65–80.
50.
FanJLinSK.Test of significance when data are curves. J Am Stat Assoc1998;
93: 1007–1021.
51.
GóreckiTSmagaŁ.A comparison of tests for the one-way ANOVA problem for functional data. Comput Stat2015;
30: 987–1010.
52.
ShenQFarawayJ.An f test for linear models with functional responses. Stat Sin2004;
14: 1239–1257.
53.
StoreyJDXiaoWLeekJT, et al.
Significance analysis of time course microarray experiments.Proc Natl Acad Sci U S A2005;
102: 12837–12842.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
